Motion artifact correction in free‐breathing abdominal MRI using overlapping partial samples to recover image deformations

Abstract

This article presents a method to reconstruct liver MRI data acquired continuously during free breathing, without any external sensor or navigator measurements. When the deformations associated with k-space data are known, generalized matrix inversion reconstruction has been shown to be effective in reducing the ghosting and blurring artifacts of motion. This article describes a novel method to obtain these nonrigid deformations. A breathing model is built from a fast dynamic series: low spatial resolution images are registered and their deformations parameterized by overall superior-inferior displacement. The correct deformation for each subset of the subsequent imaging data is then found by comparing a few lines of k-space with the equivalent lines from a deformed reference image while varying the deformation over the model parameter. This procedure is known as image deformation recovery using overlapping partial samples (iDROPS). Simulations using 10 rapid dynamic studies from volunteers showed the average error in iDROPS-derived deformations within the liver to be 1.43 mm. A further four volunteers were imaged at higher spatial resolution. The complete reconstruction process using data from throughout several breathing cycles was shown to reduce blurring and ghosting in the liver. Retrospective respiratory gating was also demonstrated using the iDROPS parameterization.Magn Reson Med 62:440-449, 2009.

Respiratory motion models: A review

“…This section discusses sources of measurements of the true respiratory motion that are used, typically together with the surrogate data, to determine the correspondence model. Section 5.2 Wang et al, 1995;Blackall et al, 2001;King et al, 2001;Manke et al, 2002b;McLeish et al, 2002;Blackall et al, 2005;Blackall et al, 2006;Reyes et al, 2007 Respiratory gated images Manke et al, 2002a;Ablitt et al, 2004;Wu et al, 2006;Buerger et al, 2012;Preiswerk et al, 2012 Dynamic images Manke et al, 2003;Khamene et al, 2004;Koch et al, 2004;Liu et al, 2004;Sundaram et al, 2004;Jahnke et al, 2005;Nehrke and Bornert, 2005;Plathow et al, 2005;Blackall et al, 2006;Fischer et al, 2006;Jahnke et al, 2007;Sharif and Bresler, 2007;Gao et al, 2008;King et al, 2008a;King et al, 2008b;King et al, 2009a;King et al, 2009b;White et al, 2009;King et al, 2010b;King et al, 2010c;Rijkhorst et al, 2010;King et al, 2011;McGlashan and King, 2011;Rijkhorst et al, 2011;Savill et al, 2011;King et al, 2012;Peressutti et al, 2012…”

“…For retrospective correction, the mathematical theory for nonlinear motion was described in Batchelor et al (2005) and first applied using a motion model based technique in Odille et al (2008b). Subsequent works include Filipovic et al (2011), Odille et al (2008a, Odille et al (2010), and White et al (2009). In addition, a number of general motion models have been proposed without being specific about which motion-correction strategy they could be used for (Jahnke et al, 2007;Manke et al, 2002a;Nehrke et al, 2001;Sharif and Bresler, 2007;Wang et al, 1995).…”

“…The first description of a motion modelling technique using images as surrogate data was that of King et al (2001) using 2-D US. Note that in all cases a motion model was required because it was not feasible to estimate the motion directly from the images, either because of a limited amount of data Li et al, 2011a;Vandemeulebroucke et al, 2009;White et al, 2009) or because of poor quality images (King et al, 2010b). All of the above methods were image driven approaches, based on indirect correspondences (see Sections 1.1.1 and 5.3.2).…”

“…Proposals for using images as surrogate data have included cone beam CT projections (Li et al, 2011a;Vandemeulebroucke et al, 2009); 2-D King et al, 2001) and 3-D (King et al, 2008b(King et al, , 2010bPeressutti et al, 2012) US; or a small number of lines of k-space (White et al, 2009) acquired during MR imaging. Although 2D MR 'navigators' are gaining attention in the literature on respiratory motion correction (Henningsson et al, 2012;Keegan et al, 2007;Stehning et al, 2005;Uribe et al, 2007) to the authors' knowledge the only paper to have used such images as surrogate data for a motion model is King et al (2012).…”

Patient-initiated breath-holds in MRI: an alternative for reducing respiratory artifacts and improving image quality

“…There are a number of well-known strategies to deal with this problem including respiratory gating/triggering, and adapting the phase encoding direction to the moving structures in the region of interest [1]. Both retrospective [2,3] and prospective [4] correction methods have been proposed to address the problem of respiratory motion artifacts. The PROPELLER method [5], which combines data collection and reconstruction for motion correction, has also been shown to be valuable for regions susceptible to respiratory motion artifacts [6,7].…”